View Proposal

Proposer
Mohammad Ammad-Uddin
Title
AI and Data-Driven Computing for Energy, Climate and Sustainable Systems
Goal
To investigate, design, implement and evaluate an intelligent data-driven solution for a selected problem in energy, climate, renewable-energy systems, smart infrastructure or sustainable agriculture. Each student will identify an appropriate application domain, investigate suitable datasets and computational methods, develop a working prototype or analytical system, and critically evaluate its performance.
Description
Energy and climate systems increasingly generate large quantities of data from renewable-energy installations, smart meters, environmental sensors, weather services, IoT systems and agricultural monitoring platforms. Artificial intelligence, machine learning, data analytics, simulation and optimization can be used to analyse these data and support forecasting, monitoring, decision-making and efficient resource management. This project will investigate the application of intelligent computing techniques to a selected problem within the broad area of energy, climate and sustainable systems. The student will first investigate potential application areas and select a suitable problem based on its practical relevance, availability of data, computational requirements, research potential and feasibility within the project timeframe. Depending on the selected application, the student may work with variables such as renewable-energy generation, solar irradiance, temperature, cloud cover, wind conditions, energy consumption, battery state, soil moisture, crop or environmental conditions, sensor measurements or other relevant time-series and contextual data. The student will identify and obtain suitable public datasets or other appropriately available data sources. The data should be prepared through appropriate processes such as cleaning, missing-value treatment, synchronization, exploratory analysis and feature engineering. Suitable computational approaches will then be investigated and selected. These may include machine-learning regression or classification, time-series forecasting, anomaly or fault detection, optimization, simulation, decision-support methods, explainable artificial intelligence, or hybrid data-driven and model-based approaches. The student should justify the selected methodology rather than being given a predefined model or algorithm. Where appropriate, more than one approach should be implemented and compared with a suitable baseline. A working prototype, analytical pipeline, simulation or decision-support system should be developed for the selected application. Evaluation will depend on the chosen problem. Suitable measures may include prediction error, classification accuracy, precision, recall, F1-score, anomaly-detection performance, computational performance, energy efficiency, resource utilization, robustness, optimization objectives or other domain-relevant measures. The student should also investigate how the developed system behaves under changing conditions. Examples include different seasons, weather patterns, energy-demand levels, renewable-energy availability, extreme-weather conditions or changes in environmental parameters. The final project should demonstrate the student's ability to identify and define an appropriate energy, climate or sustainability problem; investigate suitable datasets and computational methods; design and implement an intelligent solution; conduct appropriate experiments; critically evaluate the resulting system; and discuss practical limitations and opportunities for future improvement. Students allocated to this project theme must undertake sufficiently different applications or research questions so that each project has an independent technical contribution, implementation and evaluation. Example Application Domains (Illustrative Only) Example 1 - Solar Energy Forecasting A student may investigate forecasting of photovoltaic electricity production using historical PV generation together with weather variables such as solar irradiance, temperature, cloud cover and wind conditions. The project may compare different forecasting techniques and evaluate their performance for next-hour, day-ahead or other forecasting horizons under different seasons and weather conditions. Example 2 - Smart Energy Consumption and Anomaly Analysis A student may investigate energy-consumption prediction or anomaly detection in smart buildings, photovoltaic installations or other energy-IoT systems. The student could develop models that learn expected energy behaviour and identify unusual consumption patterns, equipment faults, sensor problems or unexpected reductions in renewable-energy production. Example 3 - Agrivoltaic Solar-Crop Optimization A student may investigate how photovoltaic electricity generation can be balanced against the sunlight and shading requirements of crops in an agrivoltaic system. For example, crops may require a specified minimum period of direct sunlight and a suitable period of shade during the day. The student could model solar position, panel configuration and shadow patterns and estimate photovoltaic electricity generation under different panel orientations, tilt angles, spacing or tracking strategies. The objective would be to identify configurations that provide suitable light conditions for plants while maintaining effective renewable-energy generation. The project could therefore investigate the trade-off between crop sunlight requirements, crop shading requirements and photovoltaic energy production.
Resources
Energy4Climate (E4C) Living Lab / agrivoltaic context: https://www.e4c.ip-paris.fr/ Energy4Climate SIRTA-PV1 public data: https://gitlab.in2p3.fr/energy4climate/public/sirta-pv1-data NASA POWER agroclimatology and solar data: https://power.larc.nasa.gov/ PVGIS: https://joint-research-centre.ec.europa.eu/pvgis-online-tool_en FAO AquaCrop: https://www.fao.org/aquacrop
Background
Url
Difficulty Level
High
Ethical Approval
None
Number Of Students
3
Supervisor
Mohammad Ammad-Uddin
Keywords
agrivoltaics, smart agriculture, solar energy, iot, machine learning, optimisation, digital twin, irrigation, microclimate, water-energy-food nexus
Degrees
Bachelor of Science in Computational Sciences and Software Engineering
Back to List